Thanks to brave testers in #lighttpd the AIO-support is stabilizing very well and the corruptions that have been reported are fixed now.

Next to bugfixes, I implemented chunk-stealing and doubled the performance of aio for small files (100k) [16MByte/s instead of 9MByte/s].

Download: www.lighttpd.net/download/lighttpd-1.5.0-r1454.tar.gz

Yeah, really.

Before you jump around and empty a barrel of beer, try to compile it first. :)

Download: www.lighttpd.net/download/lighttpd-1.5.0-r1435.tar.gz

mod-proxy-core has 3 different balancers for different needs. Round Robin, Shortest Queue First and CARP.

The benchmarks only showed results for small files (100kbyte). Time to add larger files to the pool and talk about the chunk-size.

1.5.0 will be a big win for all users. It will be more flexible in the handling and will have huge improvement for static files thanks to async io.

The following benchmarks shows a increase of 80% for the new linux-aio-sendfile backend compared the classic linux-sendfile one.

trunk/ just got support Linux Native AIO.

You might wonder why it takes to long to release 1.5.0 when most of it is already in trunk.

At MySQL we are in the final strokes of getting a GA release of Monitoring and Advisoring Service of MySQL Enterprise out of the door.

I’m still monitoring the IRC channel, but all development time is going into my MySQL stuff right now.

Only 2 weeks after .12 hit the servers we have a new release cleaning up
the issues that were introduced by it.

On the fix side we have:

• fixed a seg-fault in the HTTP-Request splitting
• fixed long-standing bug with Content-Length and HEAD requests
• fixed a possible abort of a upload if xattr is enabled

New are

• mod-magnet finally handles ‘require “lfs”’ without complaining
• mod-magnet got light.stat() which uses the stat-cache
• mod-webdav supports LOCK if compiled with —with-webdav-locks

Debian user have to compile their lua-support with:

$ configure —with-lua=lua5.1 …

Enjoy this release and watch out for 1.5.0 on the horizon. :)

Download

• lighttpd-1.4.13-1.i386.rpm [built on Fedora Core 4]
  MD5: 3d4a857e02e111d6955ccf76e416cb41
• lighttpd-1.4.13-1.src.rpm
  MD5: 6272d8310fae8bcc35e6ab4778e2016c
• lighttpd-1.4.13.tar.gz
  MD5: d775d6478391b95d841a1018c8db0b95

Back in the times of the first implementations of mod-cml I took the request setup costs as the root of all evil. They were the problem I wanted to fix with mod-cml.

But what is the request setup cost ? What is influencing the request-time ? Where can you influence it ?

It looks like mod_magnet gets more and more attraction.

• Paul Querna thinks about a magnet-like module for apache
• On marsorange is step by step install for mod-magnet on OS X
• darix finally got all the pieces together for Dr Magneto vs Mr 404 handler

Against the least pre-release we have some minor bug-fixes and the new lighty.stat() function for mod-magnet which is using our internal stat-cache to reduce the number of stat() calls which hit the kernel. darix had some nice benchmarks on it.

Download: www.lighttpd.net/download/lighttpd-1.4.13-r1385.tar.gz